Archive · 19 Dec 2025
OpenAI Launches GPT-5.2 Codex for Smarter Coding
OpenAI has launched a new AI model called GPT-5.2 Codex.
This model is special because it helps with coding tasks like writing new code, fixing bugs, and managing large codebases.
It can work on tasks for several hours without losing focus and can handle multiple steps in a single job.
OpenAI says this model is better at understanding and completing complex coding tasks compared to earlier versions.
It is available through ChatGPT, command-line tools, and integrated development environments (IDEs).
The company believes AI tools like Codex can assist developers by automating parts of the coding process, but humans will still be responsible for reviewing and approving the code.
Future updates will focus on making the model even faster and safer to use in real-world systems.
OpenAI introduced GPT-5.2 Codex, an AI model designed for coding tasks.
The model improves on writing, understanding, and managing code over long and complex projects.
GPT-5.2 Codex can handle tasks like writing new code, fixing bugs, and reviewing pull requests.
It scored 55.6% on SWE-Bench Pro, a benchmark for real software engineering problems.
The model is available through ChatGPT, command-line tools, and IDEs.
- Who
- OpenAI
- What
- Launched GPT-5.2 Codex, an AI model for coding tasks
- When
- Not specified
- Where
- Announced on OpenAI’s official website
- Why
- To improve AI-assisted software development and assist developers in coding tasks
- Model Name
- GPT-5.2 Codex
- Primary Function
- AI coding agent
- Key Improvements
- Better performance on complex coding benchmarks, improved reasoning, reduced coding errors
- SWE-Bench Pro Score
- 55.6%
- Availability
- ChatGPT, command-line tools, IDEs
- Future Focus
- Faster task completion, better project understanding, safer deployment